Inverse modeling of methane sources and sinks using the adjoint of a global transport model

An inverse modeling method is presented to evaluate the sources and sinks of atmospheric methane. An adjoint version of a global transport model has been used to estimate these fluxes at a relatively high spatial and temporal resolution. Measurements from 34 monitoring stations and 11 locations along two ship cruises by the National Oceanographic and Atmospheric Administration have been used as input. Recent estimates of methane sources, including a number of minor ones, have been used as a priori constraints. For the target period 1993–1995 our inversion reduces the a priori assumed global methane emissions of 528 to 505 Tg(CH4) yr−1 a posteriori. Further, the relative contribution of the Northern Hemispheric sources decreases from 77% a priori to 67% a posteriori. In addition to making the emission estimate more consistent with the measurements, the inversion helps to reduce the uncertainties in the sources. Uncertainty reductions vary from 75% on the global scale to ∼1% on the grid-scale (8x10), indicating that the grid scale variability is not resolved by the measurements. Large scale features such as the interhemispheric methane concentration gradient are relatively well resolved and therefore impose strong constraints on the estimated fluxes. The capability of the model to reproduce this gradient is critically dependent on the accuracy at which the interhemispheric tracer exchange and the large-scale hydroxyl radical distribution are represented. As a consequence, the inversion-derived emission estimates are sensitive to errors in the transport model and the calculated hydroxyl radical distribution. In fact, a considerable contribution of these model errors cannot be ignored. This underscores that source quantification by inverse modeling is limited by the extent to which the rate of interhemispheric transport and the hydroxyl radical distribution can be validated. We show that the use of temporal and spatial correlations of emissions may significantly improve our results; however, at present the experimental support for such correlations is lacking. Our results further indicate that uncertainty reductions reported in previous inverse studies of methane have been overestimated.